Airbnb. has been granted a patent for a method that identifies correlations between property listings from different merchant platforms. The process involves comparing images of listings based on keypoints and determining similarities to ascertain if they represent the same property. GlobalData’s report on Airbnb gives a 360-degree view of the company including its patenting strategy. Buy the report here.

Smarter leaders trust GlobalData

Report-cover

Data Insights Airbnb Inc - Company Profile

Buy the Report

Data Insights

The gold standard of business intelligence.

Find out more

According to GlobalData’s company profile on Airbnb, AI for workflow management was a key innovation area identified from patents. Airbnb's grant share as of July 2024 was 85%. Grant share is based on the ratio of number of grants to total number of patents.

Method for identifying correlated property listings using image analysis

Source: United States Patent and Trademark Office (USPTO). Credit: Airbnb Inc

The patent US12067634B2 outlines a method and system for identifying correlations between property listings using image analysis and geographic data. The process begins by selecting a source property listing and a target property listing within a defined geographic area. Images associated with both listings are stored in separate databases. The method involves identifying keypoints in these images and categorizing them (e.g., kitchen, bathroom, living room). By comparing the keypoint descriptors of the source images with those of the target images, the system determines the number of shared keypoints, which informs the likelihood of correlation between the two listings. The method also considers the pairing of images and their respective categories to enhance the accuracy of correlation assessments.

Additionally, the patent specifies that the likelihood of correlation is influenced by the image categories, with a higher likelihood associated with images categorized as kitchen, bathroom, or living room. The system can also remove images categorized as views or pools to refine the analysis. Furthermore, the method incorporates a comparison of description data related to the source and target listings, adding another layer of evaluation for correlation. The geographic area for analysis can be defined as either a square or circular region, with a predetermined size of 1.3 km. Overall, the patent presents a comprehensive approach to leveraging image and geographic data for real estate analysis, potentially enhancing property listing evaluations and market insights.

To know more about GlobalData’s detailed insights on Airbnb, buy the report here.